However, language learning is, after all, learning language, not just doing fun things with it. And texts – by which I mean the relatively short spoken and written passages that come in textbooks and other teaching materials – can, if they are used properly, play an important part in the learning process. So here goes.
Three kinds of input
Let's start by looking at the overall structure of language learning. It is useful to identify three kinds of useful input: extensive, intensive and analysed. Children learning their mother tongues receive massive extensive input from the cloud of language that surrounds them, some of it roughly attuned to their level of development, much of it not. They also receive substantial intensive input – small samples of language such as nursery rhymes, stories, songs, the daily mealtime and bedtime scripts, and so on, which are repeated, assimilated, memorised, probably unconsciously analysed, and/or used as templates for future production. And children receive analysed input: explicit information about language. Although they are not generally told very much about grammar and pronunciation, they constantly demand explanations of vocabulary: ‘What’s a …?’; ‘What’s that?’; ‘What does … mean?’
Second-language learners are no different in principle from small children in these respects. They, too, need extensive input – exposure to quantities of spoken and written language, authentic or not too tidied up, for their unconscious acquisition processes to work on. (For evidence for the effectiveness of extensive reading, see for example Day and Bamford 1998, or Alan Maley's survey of the research in his December 2009 guest article.) Equally, learners need intensive engagement with small samples of language which they can internalise, process, make their own and use as bases for their own production (Cook 2000). And since most instructed second-language learners have only a fraction of the input that is available to child first-language learners, the deliberate teaching of grammatical as well as lexical regularities – analysed input – helps to compensate for the inadequacy of naturalistic exposure for at least some aspects of language.
Three kinds of output
Input is only half the story. People generally seem to learn best what they use most. Children produce quantities of extensive output, chattering away as they activate what they have taken in. They also recycle the intensive input they have received, repeating their stories, nursery rhymes and so on, and speaking their lines in the recurrent daily scripts of childhood life. And some children, at least, seem to produce certain kinds of analysed output, naming things or rehearsing and trying out variations on structures that they have been exposed to, like more formal language learners doing ‘pattern practice’ (Weir 1970).
Adults, too, need opportunities to produce all three kinds of output. They must have the chance to engage in extensive, ‘free’ speech and writing; they must be able to systematically recycle the intensive input that they have more or less internalised (and thus complete the process of internalisation); and they need to practise the analysed patterns and language items that have been presented to them, so that they have some chance of carrying them over into spontaneous fluent production.
A properly-balanced language-teaching programme, then, will have these three ingredients – extensive, intensive and analysed – at both input and output stages. While all the ingredients are important, the proportions in a given teaching programme will naturally vary according to the learners' needs, their level, and the availability of each element both in and out of class.
What can texts do?
So where do textbook texts – relatively short continuous pieces of spoken or written language – come into all this? Clearly they can contribute in various ways to the three-part process outlined above. They can provide material for practice in receptive skills, and thus facilitate access to extensive input. They can act as springboards for discussion, role play, or other kinds of extensive output work. They can support analysed input by contextualising new language items. A further role – and a very important one – is to provide the intensive input that all learners need: short samples of appropriately selected language which are carefully attended to and partly internalised, and which can then serve as a basis for controlled production.
What do texts usually do?
Unfortunately, this aspect of text use is often neglected or ineffectively put into practice. A language-teaching text may simply be seen as something to be ‘gone through’ in one way or another, without any clear definition of the outcomes envisaged. (Text-work is an awfully convenient way of filling up a language lesson, and teachers often feel that any text-based activity is bound to be beneficial. This is not necessarily the case.) One approach to ‘going through’ is the traditional pseudo-intensive lesson where the teacher uses a text as the basis for a kind of free-association fireworks display. He or she comments on one word, expression or structure after another, elicits synonyms and antonyms, pursues ideas sparked off by the text, perhaps gets the students to read aloud or translate bits, and so on and so on. Meanwhile the students write down hundreds of pieces of information in those overfilled notebooks that someone once memorably called ‘word cemeteries’. When the end of the 'lesson' is approaching, students may answer some so-called ‘comprehension questions’. (As Mario Rinvolucri asked in his November 2008 guest article, what exactly are these for? If you have spent an hour working on a text with your class and still need to find out whether they understand it, perhaps there's something wrong.). Students then go away to write a homework on a topic distantly related (or even not at all related) to that of the text. This kind of activity tends to fall between two stools: the text is too short to contribute much to learners' extensive experience of language, but the work done on it is not really intensive either. At the end of the cycle the students have been given much too much input, have engaged with it too superficially to assimilate much of it, and have used (and therefore consolidated) little or none of it. They have been taught – inefficiently – one lot of language, and then asked to produce a substantially different lot.
Another approach which has been fashionable in recent decades is to use a written text to teach 'reading skills'. The text is typically accompanied by a battery of exercises which require students to predict, skim, scan, identify main ideas, match topics to paragraphs, sort out shuffled texts, and so on. There is an implicit assumption that even perfectly competent mother-tongue readers actually need to learn to process text all over again in a new language. For a critique of this view, see Walter and Swan 2008. Here again, students may spend substantial time working through a text without any very identifiable payoff in terms of increased language knowledge or genuine skills development.
While texts can undoubtedly be valuable in various ways, I believe they are best used with a clear purpose in mind, and a reasonable certainty that they will help to achieve this purpose. In a second article I will focus on the intensive input-output cycle referred to above, which I believe is centrally important, and I will consider ways in which texts can be exploited efficiently to support this aspect of language learning.
Cook, G. 2000. Language Play, Language Learning. Oxford: Oxford University Press.
Day, R. & Bamford, J. 1998. Extensive Reading in the Second Language Classroom. Cambridge: Cambridge University Press.
Walter, C. & Swan, M. 2008. 'Teaching reading skills: mostly a waste of time?' in IATEFL 2008: Exeter Conference Selections.
Weir, R. H. 1970. Language in the Crib. The Hague: Mouton.
Read the follow up article by Michael Swan on texts here.
This article was first published in January 2011